Extracting Power Signature from Low Dropout Voltage Regulator for IoT Security

David Thompson and Haibo Wang
Southern Illinois University Carbondale


Abstract

Power analysis or power fingerprinting has been demonstrated as a potent tool to detect malicious operation or adversary attacks. However, current methods to extract power traces used in power fingerprinting typically require the use of bulky and power hungry bench equipment, which prevents integrating the power analysis method in miniature IoT devices. This work presents a novel method to extract the power signature of IoT devices. The proposed method takes advantage of the digital control signals of digital LDO circuits and generates the power signature with simple circuits. The proposed circuits can be integrated into LDO chips and hence is suitable for IoT applications. The proposed circuit was designed using a 130 nm CMOS technology and its performance is evaluated via simulation with real sensor power trace data.